Faculty of Biosciences, Fisheries and Economics School of Business and Economics
Antecedents of consumer evaluation and consumption of functional food
Theory of planned behavior, personality traits, personal values, and future time perspective Bjørn Tore Nystrand
A dissertation for the degree of Philosophiae Doctor. September 2021
i
ii Abstract
Purpose—Functional foods are food products that have been enriched with minerals, vitamins, fatty acids, or proteins to make them healthier or to prevent diseases. Functional food is an ever-growing global phenomenon and new product launches to the category occur frequently. The term functional food is, however, ambiguous and prone to misunderstandings. Little is known about Norwegian consumers’ perceptions and evaluations of functional food. To gain a deeper understanding of the psychological mechanisms involved in consumers’ evaluation of and behavioral tendencies regarding the consumption of functional food, this thesis aims to identify, explore, test, and explain whether and how behavioral intentions and consumption of functional food (behavioral tendencies) are influenced by consumers’ attitudes (cognitive and affective), social norms (descriptive and injunctive),
perceptions about behavioral control (controllability and self-efficacy), time perspective, individual personality traits, and food-specific personal values. The main objectives of this dissertation are thus (a) to test and establish whether an extension of the theory of planned behavior has increased
explanatory ability, (b) to investigate the roles of personality traits and time perspective in explaining consumption, and (c) to explore and discuss whether consumer profiles based on individual, food- related values and traits are differentially related to consumer evaluation and consumption of
functional food. The current research is guided by the theory of planned behavior, which is one of the most frequently used social cognition models employed to identify, explain, and predict health and food consumption behaviors.
Design/methodology/approach—The papers are based on survey data from a representative sample of 810 Norwegian adults. Structural equation modeling (SEM), using AMOS and RStudio with the lavaan package, was applied to analyze the data in paper 1 and paper 2, respectively. For paper 3, hierarchical k-means clustering, using the packages cluster and factoextra in RStudio, was applied.
Paper 1 addresses the antecedents of consumers’ attitudes and behavioral tendencies toward the consumption of functional food using an extended version of the theory of planned behavior. In paper 2, an integrative and hierarchical structure of personality traits–food-related time perspective–
consumption behavior is specified and tested. Finally, paper 3 explores the intraindividual
organization of food- and health-related values and traits to determine the existence of subgroups of consumers with similar profiles, and whether attitude, intention, and consumption behavior
differentiate subgroups.
Findings and contributions—The first paper addressed antecedents of consumers’ attitude and
intention to consume functional food using an extended version of the theory of planned behavior. The results demonstrated that the extended model increased explained variance from 64.8% to 70.5% (ƒ2 = 0.19; medium- to large-effect size). Self-efficacy shared the strongest association with intention to
iii
consume functional food, followed by attitude and social norms. Except for controllability, which was negatively related to intention, all of the other antecedents of intention had the expected positive sign.
Utilitarian eating value contributed strongly and positively to attitude formation, while the relationship between hedonic eating value and attitude was weak and negative. The increased explained variance provides a deeper understanding of consumers’ motivation and behavior related to the consumption of functional food. The occasionally used self-efficacy-as-motivation argument was here rejected as an explanation for the strong association between self-efficacy and intention.
The purpose of paper 2 is twofold: First, it contributes to the ongoing debate on time perspective dimensionality and specifies and compares a unidimensional to a bidimensional measurement model of consideration of future consequences (CFC) pertaining to food choices. Second, it addresses how the big five personality traits and CFC interact in explaining variation in functional food consumption.
The results are supportive of a bidimensional factor structure for CFC, wherein one dimension reflects consideration of immediate consequences (CFC-Immediate) and the other taps into consideration of future consequences (CFC-Future). Both CFC-Future and CFC-Immediate are positively associated with functional food consumption, the former being the stronger predictor. A comparison between the two main models—a full vs. partial mediation model—yields support for retaining the partial
mediation model. It suggests that the personality traits conscientiousness, agreeableness, and neuroticism are positively and indirectly associated with consumption of functional foods through CFC-Future.
Paper 3 sought to explore and determine the existence of subgroups of consumers with similar profiles or combinations of traits and values and to investigate how attitudes and behavioral tendencies toward eating functional food differ between these consumer segments. The results demonstrate how food- and health-related values and traits can successfully discriminate between homogeneous groups of consumers to form useful consumer profiles. The person-centered approach allows for uncovering and understanding consumer profiles based on combinations of food-relevant personality traits and personal values. The three consumer profiles identified—convenience-oriented, self-controlled, and careless—differ in their propensity to consume functional foods and their attitudes and intentions regarding functional food consumption. The combination of being convenience oriented, having a low level of self-control, and being concerned about weight gain is a significant descriptor of consumers more inclined to favor functional food.
This thesis contributes to the existing literature on the explanation and understanding of individual differences in the evaluation and consumption of functional food by combining a variable- or construct-centered approach (SEM) with a person-centered approach (clustering technique). The
iv
present research enhances the understanding of the underlying motivations behind consumers’
evaluation of and behavior toward functional foods. One key finding from paper 1 is the differential influence that hedonic and utilitarian eating values exert on attitude toward consuming functional food. From paper 2, the differential effect of future- and immediate time perspectives on consumption of functional food represents one important contribution. Another key finding is the nature of the personality trait–time perspective–behavior relationship. Finally, in addressing the intraindividual organization of food-related personality traits and personal values, paper 3 confirms how consumer profiles differ in their evaluations of and propensities to consume functional food.
Research limitations—The current research relies on a cross-sectional design with self-reported data, which entails drawbacks, particularly regarding causal explanation and method biases. The
retrospective measure of consumption frequency also renders causal explanations inappropriate.
Another general limitation lies in the superordinate definition of functional food employed—that is, disregarding specific functional food products and rather focusing on functional food as an
overarching food category. Alternative research designs are welcomed and plans for conducting consumer and sensory experiments have been initiated.
Originality/value—Functional food still has limited outreach in the Norwegian marketplace compared with other markets. This research is, to the author’s knowledge, the first comprehensive effort to gain important insights into consumers’ evaluations of and motivations toward consuming foods enriched with functional components. The results should therefore be useful to better target functional food according to consumers’ motivational antecedents and personal features.
v
vi Acknowledgements
This dissertation marks the end of a long and winding road. It all began while on paternity leave following the birth of Pia, my firstborn daughter. Møre og Romsdal County municipality had only just agreed to fund my project when it was time for me to pick up the baton from my wife and stay home.
When I returned from paternity leave, all the formalities were in place, and I could start my PhD journey at the School of Business and Economics, UiT The Arctic University of Norway in late 2017.
Although I have been formally affiliated with Tromsø, throughout this period I have been physically located in Ålesund where I live and where my workplace at Møreforsking is. I will admit that the spatial distance between the two cities and between my supervisor, Professor Svein Ottar Olsen, and myself has been challenging from time to time. Svein Ottar’s commitment, guidance, and willingness to share from his reservoir of expertise during these years, however, outweighs the distance and far exceeds my hopes and expectations. I could not have done this without you, and I am very grateful that you wanted to be my mentor and supervisor, Svein Ottar.
Before turning my attention to other significant supporters, I feel obligated to mention a life-changing event that took place as I entered what was to be the final year of my PhD. My wife and I were expecting our long-awaited second child, the last piece of the family puzzle. Unfortunately, Håkon was terminally ill when he was born and never got to experience the outside of the neonatal intensive care unit where he died only 20 days old, March 20, 2020. Our stay in the hospital coincided with the onset of the coronavirus pandemic, which imposed a ban on hospital visits and effectively deprived Pia of the opportunity to even get to meet her baby brother. The reconciliation of his passing is persistent and at times all-consuming. How we wish you were here, Håkon.
I would like to express appreciation to Møreforsking for giving me the opportunity and time to pursue a doctorate degree in my current position at the institute. I would also like to acknowledge Møre og Romsdal County municipality for financial support.
Finally, I wish to express my sincerest gratitude to my family, and most importantly my wife, Anniken, who for some reason unbeknownst to me has remained confident that I would be able to complete this project. Thank you for your enduring support—I am forever grateful. And not to forget Pia, who never ceases to amaze us, always happy, and, without a doubt, our most significant
achievement in life.
Bjørn Tore Nystrand Ålesund, September 2021
vii
viii Table of Contents
PART 1. INTRODUCTION ... 1
1.1 Background ... 1
1.2 Exploring antecedents of functional food consumption ... 3
1.3 Exploring antecedents of consumer evaluation and consumption of food ... 5
1.4 Exploring gaps in explaining evaluation and consumption of functional food ... 10
1.5 Overall conceptual framework for this thesis ... 13
1.6 Extending the theory of planned behavior ... 14
1.6.1 Behavioral tendencies: Intention and consumption ... 14
1.6.2 Evaluation: Cognitive and affective attitudes ... 16
1.6.3 Social norms: Injunctive and descriptive ... 17
1.6.4 Perceived behavioral control: Controllability and self-efficacy ... 19
1.7 Individual differences ... 21
1.7.1 Personality traits ... 21
1.7.2 Future time perspective: CFC ... 25
1.7.3 Domain-specific personal values ... 29
1.8 Research design and data ... 33
1.8.1 Survey design ... 34
1.8.2 Sample and procedure ... 36
1.8.3 Measures and construct validation ... 40
PART 2. MAIN FINDINGS, DISCUSSION, AND CONTRIBUTIONS ... 45
2.1 Extending the TPB to better explain intention and consumption behavior ... 46
2.2 The roles of personality traits and time perspectives ... 48
2.3 Profiling consumers based on traits and values ... 50
2.4 Limitations and suggestions for future research ... 53
References ... 59
PART 3. PAPERS ... 83
ix PART 3. PAPERS
Paper 1. Nystrand, B. T., & Olsen, S. O. (2020). Consumers’ attitudes and intentions toward consuming functional foods in Norway. Food Quality and Preference, 80, 103827.
https://doi.org/10.1016/j.foodqual.2019.103827 Journal metrics (2020):
• Impact Factor: 5.565 (18/144 in Food Science & Technology)
• CiteScore: 8.4 (20/310 in Food Science)
• Publisher: Elsevier
• Acceptance rate: 20%
Article metrics (per September 2021):
• Citations (Google Scholar): 54
Paper 2. Nystrand, B. T., Olsen, S. O., & Tudoran, A. A. (2021). Individual differences in functional food consumption: The role of time perspective and the Big Five
personality traits. Appetite, 156, 104979. https://doi.org/10.1016/j.appet.2020.104979 Journal metrics (2020):
• Impact Factor: 3.868 (ranks 8/53 in Behavioral Sciences)
• CiteScore: 7.0 (ranks 14/203 in General Psychology)
• Publisher: Elsevier
• Acceptance rate: 19%
Article metrics (per September 2021):
• Citations (Google Scholar): 3
Paper 3. Nystrand, B. T., & Olsen, S. O. (2021). Relationships between functional food consumption and individual traits and values: A segmentation approach. Journal of Functional Foods, 86, 104736. https://doi.org/10.1016/j.jff.2021.104736
Journal metrics (2020):
• Impact Factor: 4.451 (ranks 35/144 in Food Science & Technology)
• CiteScore: 6.8 (ranks 34/310 in Food Science)
• Publisher: Elsevier
• Acceptance rate: N/A
1
PART 1. INTRODUCTION
The overall purpose of this dissertation is to gain a deeper understanding of the psychological
mechanisms involved in consumers’ behavior regarding the consumption of functional food. This will be pursued by identifying, exploring, testing, and explaining whether and how behavioral intentions regarding, and consumption of functional food (behavioral tendencies) are influenced by consumers’
attitudes (cognitive and affective), social norms (descriptive and injunctive), perceptions about behavioral control (controllability and self-efficacy), time perspective, individual personality traits, and food-specific personal values. This study also contributes to the existing literature on the explanation and understanding of individual differences in the evaluation and consumption of functional food by combining a variable- or construct-centered approach (structural equation modeling, SEM) with a person-centered approach (clustering technique). Finally, the theory-based assumptions are tested on a nationwide and representative sample of Norwegian consumers.
1.1 Background
Eating is among the most frequent behaviors in which humans engage, thus driving great interest in understanding the underlying factors that influence food choice decisions and consumption behavior (Köster, 2009; Symmank et al., 2017). Diet is today more than just the provision of sufficient amounts of nutrients—it may also “modulate various functions in the body and may play detrimental or
beneficial roles in some diseases” (Roberfroid, 2000, p. 1660S). Increased knowledge of the dietary influence on health and well-being coupled with rising healthcare costs, longer life expectancy, an aging and ever-growing population, and scientific and technological advances, has paved the way for the concept of functional food (Kaur & Das, 2011; Roberfroid, 2000; Vergari et al., 2010). Other concurrent, and to some extent overlapping, food trends include those related to organic food (Rana &
Paul, 2017), natural food (Román et al., 2017), convenience food (Jackson & Viehoff, 2016), local food (Feldmann & Hamm, 2015), and traditional food (Verbeke, 2013).
The term functional food was coined in 1984, in Japan, following the instigation of a large-scale research program with the objective of identifying foods and food substances with disease-preventing properties (Arai, 1996; Iwatani & Yamamoto, 2019). In 1991, the Food for Specified Health Uses (FOSHU) was established as a legislative framework to ascertain that the food’s safety and
effectiveness for health fulfills the requirements needed. Other countries soon followed, and functional food is today recognized as a global phenomenon (Vergari et al., 2010).
2
Functional foods are “fuzzy” borderline food products positioned between medicine and food, promising to provide additional health benefits over conventional products (Khedkar et al., 2017;
Neupane et al., 2021; Roberfroid, 1999). They offer a combination of health and convenience (Grunert, 2010) and are recognized as part of an emerging trend in the food industry (Corbo et al., 2014; Santeramo et al., 2018). The global market for functional food is expected to reach USD 275.77 billion by 2025 (Grand View Research, 2019). Generally, functional foods are those food items providing specific health benefits beyond basic nutrition (Alongi & Anese, 2021), promoting optimal health, and reducing the risk of non-communicable diseases (Granato et al., 2020; Granato et al., 2017). Functional food products are omnipresent across food categories, particularly within the dairy, confectionery, soft drink, bakery, and baby food categories (Guiné et al., 2020; Vergari et al., 2010).
Numerous definitions of functional food have been proposed (Doyon & Labrecque, 2008; Gur et al., 2018; Kaur & Das, 2011). As of yet, no universal agreed-upon definition exists, and the concept of functional food continues to be ambiguous, often misunderstood, and abused (Topolska et al., 2021).
In fact, Roberfroid (2002) argued almost 20 years ago that no simple, universally accepted definition will ever come to exist because of the large variety of (known and yet unknown) components affecting bodily functions. To complicate matters even further, a related concept, nutraceuticals, is often used in parallel with functional food by both consumers and industry (Gul et al., 2016). Nutraceuticals do, however, differ from functional foods in that this term denotes “health-promoting compounds or products that have been isolated or purified [emphasis added] from food sources” (Aluko, 2012, p.
viii) to be “supplied in other than a food form” (Laparra & Sanz, 2010, p. 220). Functional food is thus food, while nutraceuticals are not: “nutraceuticals are commodities derived from foods, but are used in the medicinal form of pills, capsules, potions and liquids” (Shahidi, 2009, p. 376).
A common way to differentiate between types of functional food is to distinguish between fortified, enriched, and altered products on the one hand, and enhanced commodities on the other (Siró et al., 2008). Briefly, following Spence (2006), fortification implies increasing the content of existing nutrients such as adding vitamin C to orange juice. Enrichment is adding new nutrients or functional ingredients not normally found in a particular food, for example by adding omega-3 to dairy products.
Alteration involves replacing existing, potentially harmful components with other, more beneficial components. Reducing salt by replacing sodium chloride with potassium chloride or other flavor enhancers is a widely used approach to alteration. Lastly, enhanced commodities are achieved by altering the nutrient composition of raw commodities, a well-known exemplar being golden rice—an engineered variety of rice containing vitamin A (Tang et al., 2009). Other definitions of functional food also include whole foods or natural commodities such as fruits and vegetables (e.g., Martirosyan
& Singh, 2015).
3
The attractiveness of functional foods resides in combining convenience with health. Staple foods such as dairy and cereal products are made healthier by enrichment, thus providing an easy avenue for people to choose the healthier option without completely changing their consumption patterns, such as by opting for milk with added vitamin D rather than keeping with the conventional counterpart. A major barrier, on the other hand, can be ascribed to the price premiums associated with functional foods (Annunziata & Vecchio, 2011; Siró et al., 2008): The price of food is consistently one of the most important food choice motives reported (Markovina et al., 2015). Functional food has been posited to constitute the perfect marketing strategy: “creating differentiated, value-added products, appealing to health (a basic and universal human need) and directed to a premium-price sector”
(Falguera et al., 2012, p. 276). One hurdle in reviewing the literature stems from the above-mentioned confusion about what constitutes a functional food. Some (review) studies adhere to a broad definition while others to more restricted definitions (Mogendi et al., 2016). The same applies to the various estimates of market share, value, and outreach.
This dissertation adheres to the fortified/enriched products category and leans on a definition of functional food (Doyon & Labrecque, 2008; Laros & Steenkamp, 2005) that excludes naturally healthy foods such as vegetables and fruits. It specifically introduces a definition that states that functional foods are food products that have been enriched with minerals, vitamins, fatty acids, or proteins to make them healthier or to prevent diseases. Functional foods are further part of a standard diet, consumed on a regular basis, and in normal quantities. As such, dietary supplements are also excluded. The purpose of opting for such a restricted definition is to avoid any ambiguity when evaluating items related to functional food behavior. This constrained definition holds that a functional food is one in which functional ingredients have been incorporated—that is, they are a manufactured rather than a natural food product.
This thesis focuses on consumer evaluation of functional food consumption, and its main contributions lie in the identification of key psychological antecedents and individual differences that influence subjective perceptions, attitudes, and behavioral tendencies toward the consumption of functional food.
1.2 Exploring antecedents of functional food consumption
Research aimed at explaining or predicting consumer evaluation of functional food ranges from focusing on sociodemographic variables and product attributes to examining the psychological antecedents underlying behavior (for reviews, see Bimbo et al., 2017; Kaur & Singh, 2017; Mogendi
4
et al., 2016; Santeramo et al., 2018; Siró et al., 2008).1 Primary drivers include the general promotion of health and wellness, as well as disease prevention or minimization (e.g., health motivation; Siegrist et al., 2015; Thompson & Moughan, 2008). Functional foods are thus often accompanied by (front-of- package) health claims that “typically promise specific improvements in physiological functions or reduced risks of diseases” (Lähteenmäki, 2011, p. 109). Claiming that a functional food provides health benefits can, however, have negative effects on consumer perceptions and evaluations of other important product attributes, including taste and naturalness (Lähteenmäki, 2013; Lähteenmäki et al., 2010). Although health motivation is perhaps the most important antecedent of functional food consumption (Topolska et al., 2021), sensory attributes such as (good) taste should not be neglected (Siró et al., 2008), as it is unlikely that consumers will compromise on taste at the cost of health benefits (Verbeke, 2006). While findings are mixed and contradictory, psychological or cognitive antecedents such as attitude, perceptions, and beliefs are imperative for understanding consumer evaluation of functional food (Bimbo et al., 2017; Mogendi et al., 2016; Siró et al., 2008).
Regarding sociodemographic features, a near consensus has been reached regarding gender, whereby women are consistently shown to have higher levels of acceptance and to be the more likely
consumers of functional foods than men. Likewise, regarding age, being older rather than younger seems to be a sound descriptor of functional food consumers. Level of education or income level also appear as relevant variables, such that higher levels are characteristic of functional food consumers.
However, the influence of gender, age, and education varies as a function of “both the type of functional food and its claim” (Siró et al., 2008, p. 465).
Recent reviews (Bimbo et al., 2017; Santeramo et al., 2018) have also identified various personal values or personality traits that influence the consumer evaluation of functional foods, but studies investigating whether individual differences in broader behavioral dispositions relate to functional food consumption are still scarce. With respect to food consumption behavior in general, several recent reviews (Esposito et al., 2021; Lunn et al., 2014; Machado-Oliveira et al., 2020) demonstrate important links to personality traits. For example, conscientiousness is consistently related to eating healthily, while neuroticism is associated with unhealthy eating habits. Conscientiousness is further associated with a future time perspective, and both constructs relate to health-related behaviors and beneficial outcomes, including engagement in healthy behaviors (Baird et al., 2021; Kooij et al., 2018;
Murphy & Dockray, 2018). Little research has explored the role of personality traits and time perspective in explaining functional food behavior, leaving a research gap for further investigation.
1 Kaur and Singh’s (2017) review has been retracted (for retraction notice, see Kaur & Singh, 2019).
5
Table 1 provides a list of reviews and overview articles concerning functional food, particularly from a consumer behavior perspective (the list is not exhaustive).
Table 1. Example review articles and research on functional food from a consumer and market perspective
Author(s), year Scope Publisher
Menrad, 2003 overview about the market situation for functional food in Europe
J Food Eng Siró et al., 2008 review the current functional food market
situation, future potential, and main challenges
Appetite Ozen et al., 2012 assess differences in the worldwide
consumption of functional foods
Nutr Rev Bigliardi & Galati, 2013 specific emphasis on the definition and the main
examples of functional food, and future trends
Trends Food Sci Technol Corbo et al., 2014 focus on commercially available functional
beverages
Compr Rev Food Sci Food Saf
Mogendi et al., 2016 review evidence of the underlying determinants of consumer evaluation of nutritious food
Int J Food Sci Nutr Kaur & Singh, 2017 review studies investigating various facets of
consumer behavior toward functional foods
Appetite Bimbo et al., 2017 review research on consumer acceptance of and
preferences for functional dairy products
Appetite Santeramo et al., 2018 functional food as an emerging trend in the food
industry
Food Res Int Dolgopolova & Teuber, 2018 review consumers’ valuations of foods with
healthy attributes (meta-analysis)
Appl Econ Perspect Policy
Plasek & Temesi, 2019 identify aspects that make functional food credible for consumers
Appetite Birch & Bonwick, 2019 review the drivers of consumer choice of
functional food
Int. J. Food Sci. Technol.
Bakshi et al., 2020 review consumers’ attitudes toward functional foods
Curr Top Nutraceutical Res
Granato et al., 2020 define and classify functional foods, and exemplify recent and relevant studies
Annu Rev Food Sci Technol
Guiné et al., 2020 explore the factors that determine acceptance and willingness-to-pay for food innovations
Foods Topolska et al., 2021 provide better understanding of the needs and
behavior of consumers regarding functional food
Int J Environ Res Public Health
1.3 Exploring antecedents of consumer evaluation and consumption of food
The study of food consumption behavior spans many scientific disciplines, from biology, nutrition, medicine, and health sciences via food science and technology to psychology, business research, behavioral science, marketing, sociology, and consumer research (Köster, 2009; Symmank et al., 2017). A core question occupying research into food choice behavior is: “Why does who eat what,
6
when, and where?” (Köster, 2009, p. 70). Several different theoretical approaches and models have been developed to answer that question and to explain the reasons and motives behind food consumption behavior. One of the first attempts to develop a summary model in this regard was Pilgrim (1957), who suggested that consumer perception is the main antecedent of food acceptance.
Consumer perception is in turn a function of three basic factors: (a) physiological effects of the food, (b) perception of sensory attributes, and (c) influences from the environment (Steenkamp, 1993).
Pilgrim’s model has since influenced several subsequent efforts to develop models of determinants of food consumption behavior (Shepherd, 1990; Sijtsema et al., 2002). For an introduction to the early models, see Shepherd and Sparks (1994).
Many of these efforts attempt to integrate the different antecedents and motivations in interdisciplinary conceptual frameworks (e.g., Furst et al., 1996; Köster, 2009; Rozin, 2006). More recently, Stok et al.
(2017) introduced what may be the most comprehensive effort on structuring food choice determinants and influencing factors to date, namely the DONE2 framework. A recent interdisciplinary review using the DONE framework (Symmank et al., 2017) has demonstrated the vast amount of research on predictors of food decision-making available across a multitude of research disciplines, with
individual, psychological predictors having enjoyed the greatest research interest. Another recent review of the many existing conceptual models addressing the antecedents of food choice (Chen &
Antonelli, 2020) identifies three main categories, namely food-related features, individual differences, and society-related features. The resemblance to Steenkamp (1993) and others’ taxonomies of
properties of the food, person-related factors, and environment, respectively, is apparent. In a similar vein, attempts to synthesize and classify antecedents of functional food behavior have recently been made (Bimbo et al., 2017; Kaur & Singh, 2017; Mogendi et al., 2016). Kaur and Singh (2017) deviated from the tripartite classification of Chen and Antonelli (2020) and Steenkamp (1993) by introducing a fourth category to disentangle personal- from psychological factors. Mogendi et al.
(2016) also categorize antecedents into four groups: (a) nutritional knowledge and information, (b) attitude, perceptions, and consumer behavior, (c) price and product characteristics, and (d) socio- demographics. Bimbo et al. (2017) distinguish between only two broad categories: consumer- and product-related characteristics. The core constituents of the categories do, however, overlap to a large extent, and although each model brings new perspectives, the tripartite partitioning into (a) properties with the food product or object, (b) properties with the person engaged in the decision making and consumption process, and (c) external and environmental factors, seems to provide the consistent overarching main levels (e.g., Chen & Antonelli, 2020; Köster, 2009). Table 2 summarizes some of
2 Determinants Of Nutrition and Eating.
7
the most influential models or reviews of food choice and behavior and assigns the antecedents to these three overarching categories. Worth noting is that the boundaries between the three levels are still fuzzy, as pointed out by Steenkamp (1993) almost 30 years ago. For example, Stok et al. (2017) categorize extrinsic and intrinsic product attributes under environmental, micro-level factors, while Chen and Antonelli (2020) group social norms under food-related features.
Table 2. Summary of interdisciplinary models of determinants of food choice and behavior Author(s),
year
Food-related Person-related Environment-related
Pilgrim, 1957 Sensation derived from sensory attributes
Physiological factors and attitudes
Environmental influences and learning effects Khan &
Hackler, 1981
Extrinsic and intrinsic factors
Biological, physiological, and psychological, and personal factors and demographic
Situational, advertising, and season, and cultural, religious, and regional factors
Randall &
Sanjur, 1981
Extrinsic and intrinsic factors
Demographic, knowledge/skills, and attitude
Season, urbanization, and family size and composition Shepherd,
1985
Physical/chemical
properties, nutrient content
Perceptions and psychological factors
Social/cultural, availability, price, and brand
Steenkamp, 1993
Physical form, chemical/nutritional composition, sensory perception
Biological and psychological factors, and personality
Sociocultural, economic, and marketing factors
Sijtsema et al., 2002
Product characteristics, production system
Demographic, physiological, and psychological factors, attitudes
Family and society
characteristics, consumption moment, time, and place Köster, 2009 Extrinsic and intrinsic
factors
Biological, physiological, and psychological factors
Situational and socio-cultural Mogendi et
al., 2016
Extrinsic and intrinsic factors
Socio-demographic, cognitive, and behavioral factors, and knowledge Bimbo et al.,
2017
Extrinsic and intrinsic factors
Socio-demographic,
knowledge, and lifestyle, and psychological factors Kaur &
Singh, 2017
Extrinsic and intrinsic factors
Psychological and personal factors (e.g., knowledge, socio-demographics)
Cultural and social factors (e.g., cultural, and social norms, family composition) Stok et al.,
2017
Extrinsic and intrinsic factors
Biological, demographic, psychological, and situational factors
Social and cultural factors, meso-macro environmental factors (e.g., availability, price)
Chen &
Antonelli, 2020
Extrinsic and intrinsic factors
Personal-state and cognitive factors (e.g., biology, physiology, and psychology)
Sociocultural factor (e.g., cultural, economic, and political), social and physical environment
Note. The models’ different levels or groups of antecedents are rearranged to match a tripartite categorization into food-, person-, and environment-related factors. The category extrinsic and intrinsic factors denotes various properties of the food.
8
Some theories or approaches emphasize properties with the product such as the total food quality model (Grunert et al., 1996) and cue utilization of quality perception processes (Olson & Jacoby, 1972; Steenkamp, 1990) or contextual or environmental factors (Meiselman, 2006; Wansink, 2004).
Others highlight food-related lifestyles and values (Brunsø et al., 2004a), goals or goal conflicts (Stroebe et al., 2008), mood and emotions (Gibson, 2006; Köster & Mojet, 2015), social norms (Higgs, 2015; Higgs & Thomas, 2016), identity (Conner & Armitage, 1998; Terry et al., 1999), knowledge (Wardle et al., 2000), and habit strength (Verhoeven et al., 2012). More recent work has begun to explore the role of personality traits (Esposito et al., 2021), including impulsiveness or impulse buying tendency (Guerrieri et al., 2007; Verplanken et al., 2005), self-control (Hankonen et al., 2013; Salmon et al., 2014), temporal self-regulation and time perspectives (Dassen et al., 2015;
Hall & Fong, 2007), regulatory focus (Pula et al., 2014), and conscious vs. automatic patterns (Bublitz et al., 2010).
Several models or theories have been designed to explain, predict, or change health behavior,
including functional food behavior. Health behavior denotes “those personal attributes such as beliefs, expectations, motives, values, and other cognitive elements; personality characteristics; including affective and emotional states and traits; and overt behavior patterns, actions, and habits that relate to health maintenance and wellness, to health restoration, and to health improvement” (Gochman, 1982, p. 169). The core cognitive antecedents of health behavior include attitudes, norms, self-efficacy or perceived behavioral control (PBC), and self-representations (Abraham et al., 1998), many of which are integrated into social cognition models.
Social cognition models represent one group of health behavior models “used to help understand, predict and change health-relevant [behaviors]” (Conner & Norman, 1998, p. 179). Social cognition is understood as individuals’ sense-making of social situations, the assumption being that “social
[behavior] is best understood as a function of people’s perceptions [emphasis added] of reality, rather than as a function of an objective description of the stimulus environment” (Conner & Norman, 2015a, p. 7). Influential exemplars include the health belief model (HBM; Janz & Becker, 1984; Rosenstock, 1974), the protection motivation theory (PMT; Maddux & Rogers, 1983; Rogers, 1975), the social cognitive theory (SCT; Bandura, 1986), and the theory of planned behavior (TPB; Ajzen, 1991), as well as its predecessor, the theory of reasoned action (TRA; Fishbein & Ajzen, 1975). These are all motivational models, “designed with a view to identifying the variables that underlie health-related decisions, and to assess their ability to predict [behavior]” (Armitage & Conner, 2000, p. 174). A recent synthesis of the theories and models frequently applied to the study of health behaviors
9
concludes that the above-mentioned models are among the most frequently used theoretical
frameworks (Glanz et al., 2015). The models also share several conceptual similarities, including the role of attitudinal beliefs, beliefs regarding behavioral control or self-efficacy, and normative beliefs (Noar & Zimmerman, 2005). Of the above models, the TPB outperforms the others in predictive ability (Armitage & Conner, 2000). For that reason, the current research employs the TPB as the core theoretical framework. Detailed accounts of the other motivational models’ main constituents can be found elsewhere (Abraham et al., 1998; Armitage & Conner, 2000; Conner & Norman, 2015b).
A recent meta-analysis of research employing the TPB to predict or explain food choices (Nardi et al., 2019) demonstrates the model’s robustness in a food choice context. The TPB and constructs thereof have frequently been used as (part of the) conceptual frameworks in the study of functional food behavior, in particular behavioral beliefs and attitudinal constructs (Kaur & Singh, 2017; Mogendi et al., 2016). In the TPB, engaging in a behavior is determined by the intention to do so and perceptions about control over the enactment of the behavior. The most proximal antecedents of intention
formation are attitude toward performing the behavior, social norms related to performing the behavior, and perceived behavioral control over the enactment of the behavior (Fishbein & Ajzen, 2010/2015). The formation of attitudes, social norms, and PBC hinge upon, respectively, beliefs about the positive and negative outcomes of enacting the behavior (behavioral beliefs), beliefs about what others themselves do or expect you to do (normative beliefs), and beliefs about how easy or difficult it would be to engage in the behavior (control beliefs). In addition to the model’s main constructs, the reasoned action approach (i.e., TPB/TRA) acknowledges the role of background factors or individual difference variables responsible for the formation of beliefs, including personality traits and personal values, sociodemographic characteristics, and knowledge and information (Ajzen & Fishbein, 2005).
Within the TPB, both personality traits and personal values are considered background factors that indirectly influence intentions and behaviors through the model’s core constructs (i.e., attitude, social norms, and PBC; Ajzen et al., 2018). Background factors are believed to represent the origins of behavioral, normative, and control beliefs that ultimately reflect attitude, social norms, and PBC, respectively. Personality traits and personal values thus represent distal determinants of intention and behavior. Research employing the cognitive hierarchy model (VAB; Homer & Kahle, 1988) or Mowen’s (2000) 3-M model of motivation and personality, lends empirical support for the indirect influence of values and traits on behavior through attitudes (Conner & Abraham, 2001; Huynh &
Olsen, 2015; Kang et al., 2015; Milfont et al., 2010; Tudoran et al., 2009). Tudoran et al. (2009), for example, employed the VAB in analyzing consumers’ evaluation and intention to buy functional food.
Their study revealed that attitudes toward functional food completely mediated the relationship
10
between health value (i.e., importance and concerns about health) and attitude toward a specific functional food product.
1.4 Exploring gaps in explaining evaluation and consumption of functional food
The TPB is considered “the dominant theoretical approach to guide research on health-related [behavior]” (Sniehotta et al., 2014, p. 1) and is frequently employed to predict or explain consumer evaluation and consumption of food (Nardi et al., 2019). Aside from health behaviors, the TPB is also among the most used theoretical frameworks to explain or predict a number of other behaviors, including pro-environmental behaviors (Yuriev et al., 2020) and socially responsible consumer behaviors (Han & Stoel, 2017), to name only a few. The TPB is used because it offers great
explanatory power in a parsimonious way and is generalizable across behavioral domains. It further provides opportunities for extensions and the inclusion of background factors or external variables, including personality traits and personal values. Although functional foods are popular products around the world, less is known about Norwegian consumers’ motivation and use of such products.
Hence, this research considers the TPB framework as a relevant starting point for studying antecedents of consumer evaluation and consumption of functional food in Norway. The current research further considers temporal influences on consumption behavior and specifically argues for the relevancy of future time perspective in explaining consumption of functional food. Finally, the roles of personality traits and personal values in explaining behavioral tendencies toward functional food is addressed. In the following, arguments on how this thesis contributes to the existing TPB literature, as well as to the consumer food psychology literature, in exploring antecedents of evaluation and consumption of functional food are put forth.
The TPB is an extension of the TRA (Fishbein & Ajzen, 1975) “designed to predict and explain human behavior in specific contexts” (Ajzen, 1991, p. 181). Several studies have employed the TPB or constructs thereof to explain consumers’ behavioral tendencies toward functional foods (Nguyen et al., 2020; O’Connor & White, 2010; Patch et al., 2005; Xin & Seo, 2020). For example, Patch et al.
(2005) and Hung et al. (2016) establish a strong relationship between attitude and intention. In the studies by O’Connor and White (2010) and Nguyen et al. (2020), both attitude and social norms influence intention formation, and Xin and Seo (2020) find that attitude and PBC both influence purchase intention.
Further extensions to the original TPB model have been proposed, including alterations pertaining to the structure and operationalization of its core elements (i.e., attitude, social norms, and PBC; Ajzen et
11
al., 2018). Briefly, it is by now recognized that both affective and cognitive components should be included to the attitude construct, and that both injunctive and descriptive norms are needed to properly address the social norms construct. Similarly—although associated with somewhat more controversy—the PBC construct should incorporate measures of both self-efficacy and control.
Accordingly, the current research incorporates these considerations in the measures of attitude (Ch.
1.6.2), social norms (Ch. 1.6.3), and PBC (Ch. 1.6.4). Paper 1 also includes hedonic and utilitarian eating values as background factors hypothesized to influence attitude formation. Eating values constitute (domain-specific) motivational considerations underlying food choice and consumption behavior (Lusk & Briggeman, 2009; Tudoran et al., 2009). The assumption is that utilitarian (e.g., importance of avoiding health issues) and hedonic (e.g., importance of pleasure) eating values are differentially associated with attitude toward the consumption of functional food. Paper 3 further builds on the roles of personal values in explaining consumption behavior, introducing health importance, weight management concern, and convenience orientation, in combination with hedonic eating values, as bases for segmenting and profiling Norwegian consumers.
Personality traits and their influence on food choices and consumption have lately gained momentum in research (Gustavsen & Hegnes, 2020; Keller & Siegrist, 2015; Machado-Oliveira et al., 2020;
Pfeiler & Egloff, 2020). Some recent reviews (Esposito et al., 2021; Lunn et al., 2014) underpin the notion that personality traits are reliable predictors of dietary and health behavior patterns, and conscientiousness in particular shows consistent associations with various dietary behaviors and outcomes. Paper 2 addresses a gap in consumer research on functional food and tests the associations between the big five personality traits (i.e., openness to experience [hereafter openness],
conscientiousness, extraversion, agreeableness, and neuroticism) and consumption of functional foods.
Aside from the big five personality traits, research has also established relationships between
functional food consumption and other traits, such as food neophobia (Labrecque et al., 2006; Stratton et al., 2015), self-control (Barauskaite et al., 2018; Neupane et al., 2019), and consumer
innovativeness (Huotilainen et al., 2006). Paper 3 incorporates self-control and consumer innovativeness, domain-specific to food choice behavior, as individual difference constructs hypothesized to differentiate between consumer segments.
Another individual difference construct that has enjoyed increasing research attention in the domain of food consumption is future time perspective (Olsen & Tuu, 2021; Onwezen et al., 2016; van Beek et al., 2013). Engaging in health-promoting behaviors has long-term beneficial consequences and hence is suggested to be related to having a future time perspective (Hall et al., 2015; Sweeney & Culcea, 2017). Future time perspective has also been consistently associated with conscientiousness and proposed to act as a mediator between broader personality traits (e.g., big five) and health-related
12
behaviors and outcomes (Kooij et al., 2018). The consistent link between future time perspective and health outcomes has been demonstrated in a recent meta-analysis (Andre et al., 2018). Paper 2 considers time perspective as a mediating mechanism through which the big five personality traits influence functional food consumption behavior and tests both direct and indirect relationships between the constructs. Associations between these constructs have not, to the author’s knowledge, been established before in a functional food context.
Most social cognition models such as the TPB are predominantly variable-centered approaches to understanding social phenomena. Variable-centered approaches “assume that all individuals from a sample are drawn from a single population for which a single set of “averaged” parameters can be estimated” (Morin et al., 2018, p. 804). Complementary to the variable-centered approach, employed to investigate relationships between variables or constructs, is the person-centered approach (Fisher &
Robie, 2019; Howard & Hoffman, 2018; Morin et al., 2018). Person-centered approaches “consider the possibility that the sample might include multiple subpopulations characterized by different sets of parameters” (Morin et al., 2018, p. 804). In marketing, person-centered approaches are widely used to decompose markets into fewer, more homogenous market segments to improve the effects of different marketing strategies, such as marketing communication, product development, distributional
strategies, and pricing (Steenkamp & ter Hofstede, 2002; van Raaij & Verhallen, 1994).
Personality (trait) research employing variable-centered approaches looks to establish associations between traits and focal outcomes. Personality, however, represents the intraindividual organization of experiences and behaviors (Asendorpf, 2002), or the combination of traits within an individual. As such, the assumption underlying the person-centered approach is that the intraindividual combinations of traits, or the patterns in personality structure within individuals, partitions people into homogeneous personality types, consumer profiles, or subgroups of people (Asendorpf, 2002; Rammstedt et al., 2004). Hence, person-centered approaches constitute “a rich complement to traditional variable- centered methods” (Morin et al., 2018, p. 804). Some previous studies on functional food integrate traits, values, attitudes, habits, and other motivational or behavioral constructs to differentiate consumers in subgroups employing clustering techniques (e.g., Annunziata & Pascale, 2009; Ares &
Gámbaro, 2007; Szakály et al., 2012). The person-centered approach “is appropriate for investigating research questions and hypotheses aimed at (a) categorizing subjects into common subpopulations based on substantive variables and (b) understanding the relations of these subpopulations with predictors, correlates, or outcomes” (Howard & Hoffman, 2018, pp. 848–850). Paper 3 integrates and combines individual differences in personality traits and personal values context-specific to food consumption behavior and employs the person-centered approach (a) to determine the existence of
13
consumer profiles and (b) to establish whether and how different profiles are related to attitudes and behavioral tendencies with respect to the consumption of functional food.
1.5 Overall conceptual framework for this thesis
The overall aim of this dissertation is to improve the theoretical and empirical understanding of consumers’ behavioral tendencies toward functional food consumption. The main purpose is to provide a deeper understanding of antecedents that are relevant for explaining consumer evaluation of functional food and consumption behavior. The theoretical starting point is the TPB, a model that simultaneously represents both a comprehensive and a parsimonious structure for the study of behavioral antecedents. In general, it follows a trait/value–attitude–intention/behavior causal framework (Homer & Kahle, 1988) wherein the more abstract personality traits and personal values are held to influence attitude formation and subsequently the formation of intentions, which ultimately impact behavioral decisions.
Specifically, the following objectives are pursued:
a. To test and establish whether an extended version of the TPB that incorporates separate constructs of self-efficacy and descriptive norm, as well as hedonic and utilitarian eating values, as antecedents of attitude formation, provides superior explanatory ability over the basic TPB;
b. To investigate and establish the interrelationships between the big five personality traits and future time perspective and test their roles in explaining functional food consumption;
c. To investigate and discuss whether consumer profiles based on individual, food-related values and traits are differentially related to consumer evaluation and consumption of functional food;
d. To combine SEM and clustering techniques to achieve a better understanding of whether and how individual differences are related to the evaluation and consumption of functional foods;
and
e. To explore antecedents of functional food consumption behavior in a representative sample of Norwegian consumers.
Figure 1 depicts the conceptual model used to address the above-stated objectives. All of the associations between the constructs used in this thesis are included in the model. Briefly, paper 1 specifies and tests an extended TPB model (green-colored one-headed arrows). Paper 2 investigates
14
the roles of the big five personality traits and future time perspective in explaining functional food consumption behavior (orange-colored one-headed arrows). Finally, paper 3 combines domain- specific personality traits and personal values in segmenting consumers and tests whether the segments have different attitudes, intentions, and consumption patterns (blue-colored one-headed arrows). Worth noting is the blue-colored double-headed arrow between traits and values, which denotes a nondirectional relationship between the constructs. The point here being that paper 3 does not presume and test a causal relationship from traits to values (cf. Homer & Kahle, 1988), but rather combines the two to form the basis for segmenting consumers. The model’s constructs are presented and described in more detail in the following sections.
Figure 1. Conceptual model
1.6 Extending the theory of planned behavior
1.6.1 Behavioral tendencies: Intention and consumption
According to the TPB, behavior is guided by behavioral, normative, and control beliefs. Behavioral beliefs consider the likely positive and negative consequences of performing (or not performing) the behavior. Normative beliefs concern the expectations of others regarding performance of the behavior,
15
while control beliefs represent beliefs about the factors that may go against or further performance of the behavior (Ajzen, 2002a). These beliefs aggregate into the attitude toward performing the behavior, social norm (perceived social pressure), and PBC, respectively. The basic assumption is that the immediate antecedents to actual performing a given behavior is the intention to perform it and PBC.
The behavior in question throughout this dissertation is defined as consumption of functional foods on a regular basis. As a consequence, and in accordance with the principle of compatibility (Ajzen, 1988), all TPB constructs are operationalized with the behavioral definition in mind, for example attitude toward the consumption of functional foods on a regular basis.
On a cautionary note, the current research measured behavior with a single item that reflects past consumption of functional foods. In combination with self-reports and a cross-sectional research design, the model is better conceptualized as an explanatory rather than a predictive model. Although retrospective behavior can be a satisfactory proxy for future behavior (Jaccard & Blanton, 2005), cross-sectional data remain unsuitable for causal inferences. Past behavior can also influence future behavior, either directly or indirectly, thus altering the causal direction proposed in the TPB (Ajzen et al., 2018; Morwitz & Munz, 2021). The use of cross-sectional research designs and self-reports when employing social cognition models, such as the TPB, to predict or explain health behaviors, including food consumption behavior, is widespread, however (Nardi et al., 2019; Noar & Zimmerman, 2005).
Using single-item measures of retrospective consumption behavior is also common (e.g., Dunn et al., 2011; Goetzke et al., 2014; Olsen et al., 2007).
Intentions are “indications of how hard people are willing to try, of how much of an effort they are planning to exert, in order to perform the behavior” (Ajzen, 1991, p. 181). They operate as a motivational force that influences the likelihood of performing a given behavior (Morwitz & Munz, 2021). The underlying assumption is that people engage in intended behaviors and do not engage in unintended behaviors (Sheeran, 2002). No matter how intuitive and appealing this notion may sound, the relationship between intention and behavior, however, is not straightforward. This inconsistent relationship is called the intention–behavior gap and, according to a synthesis of relevant research on the topic, intention translates into action only one-half of the time (Sheeran & Webb, 2016). The magnitude of the intention–behavior gap for dietary behaviors, which is affected by the complex nature of food consumption (Dunn et al., 2011), is moderate (Conner et al., 2002; McEachan et al., 2011). Paper 1 hypothesized a positive relationship between intention and consumption of functional foods (retrospective behavior) and was intended to reflect the extent of planning, expectation, and intent toward the regular consumption of functional food.
16
1.6.2 Evaluation: Cognitive and affective attitudes
Attitude is “a psychological tendency that is expressed by evaluating a particular entity with some degree of favor or disfavor” (Eagly & Chaiken, 1993, p. 1). The entity, or attitude object, in question can take the form of abstract ideas, other people, physical objects, or concrete behaviors. The attitude construct has enjoyed a multitude of definitions, conceptualizations, and operationalizations since its introduction to the social sciences almost a century ago (Allport, 1935; Jung, 1923/1971). Most, if not all, attitude scholars agree that evaluation is the core of attitudes: “The study of attitudes is the study of evaluations” (Albarracin & Shavitt, 2018, p. 300). Evaluation, attitude object, and tendency are three essential features of attitude: “Evaluation refers to all classes of evaluative responding, whether overt or covert, or cognitive, affective, or behavioral”, directed toward an attitude object (Eagly &
Chaiken, 2007, p. 583). Tendency refers to the past experiences with the attitude object and denotes whether an individual is more or less likely to evaluate the attitude object in accordance with prior experiences.
Attitudes can be formed based solely on cognitive (beliefs and thoughts), affective (feelings and emotions), or behavioral (intentions and overt behavior) grounds or any combination of these evaluative aspects. The two dimensions of cognition and affect can differentially predict behavior (Lawton et al., 2009; Millar & Tesser, 1986), thus underpinning the importance of considering both in conceptualizing attitudes. Related to food attitudes, the affective dimension reflects the sensations derived from experiencing a food product (e.g., its sensory appeal) while the cognitive dimension reflects considerations of its functions (e.g., its nutritional composition; Voss et al., 2003).
An important consideration when employing the TPB is the principle of compatibility (Ajzen, 1988).
Briefly, it posits that the behavior in question should be explicitly defined in terms of its target, action, context, and time frame and that all of the other TPB constructs are defined in corresponding terms.
For example, the current research defined the behavior as the consumption (action) of functional foods (target) on a regular basis (time frame), subsequently measuring intention, attitude, social norms, and PBC in a similar manner (e.g., “My eating functional foods on a regular basis would be …”). The principle of compatibility suggests that the behavioral definition defines how the other constructs are to be operationalized (Ajzen, 2020). Relatedly, attitudes also vary in their level of specificity, from narrow to broad bandwidth, which influences their predictive accuracy. Consider for example an attitude toward buying a specific product (e.g., buying a can of Coca Cola). The attitude will likely predict the corresponding behavior (i.e., actually buying a can of Coca Cola) more accurately than a broad bandwidth attitude, such as attitude toward buying a soft drink (Ajzen, 2012).
17
Attitude is usually strongly associated with intention to consume or buy various foods (McDermott, Oliver, Simnadis, et al., 2015; McDermott, Oliver, Svenson, et al., 2015; Povey et al., 2000b), including functional foods (for a review, see Kaur & Singh, 2017). Patch et al. (2005), for example, found attitude to be the only significant predictor of intention to consume omega-3-enriched foods.
Similarly, Hung et al. (2016) found that attitude was the most important determinant for intention to purchase a new functional meat product. Szakály et al. (2019), employing a modified Munene model, demonstrated that attitude toward functional food was the best predictor of consumer willingness to pay. The list of studies supporting the significant role of attitude in predicting or explaining intention to consume or buy, or actual consumption or willingness to pay for functional foods is continuously growing. Paper 1 considers both affective and cognitive components of attitude and tests the hypothesis that attitude is positively associated with intention to consume functional food regularly.
Paper 3 rather operationalizes attitude as a global construct (e.g., negative–positive, bad–good).
1.6.3 Social norms: Injunctive and descriptive
Social or subjective norms embody the perceived social pressure or expectations of others regarding whether to engage in a particular behavior or not. They are “rules and standards that guide and constrain social behavior” (Melnyk et al., 2019, p. 6). Social norms and their impact on human behavior has enjoyed longstanding attention, but not without controversies. Cialdini et al. (1990), for example, introduce two opposing views: on the one side are advocates supporting the role of social norms in predicting and properly understanding human behavior, while on the opposite side are those that are resistant, arguing that the concept of social norms is vague and ill-suited to empirical testing.
According to Cialdini and colleagues (1991; 1990), a central explanation for why there have been so many discrepancies rests on definitional ambiguity—the term social norms can have several different meanings. Cialdini and Trost (1998) argue that social norms represent (a) general, societal
expectations for our behavior; (b) expectations of valued others (e.g., family or friends) for our behavior (i.e., injunctive norms); (c) our own expectations for our behavior (i.e., personal norms); and (d) standards that develop out of our observations of others’ behavior (i.e., descriptive norms). A recent meta-analysis (Melnyk et al., 2019) demonstrates the differential effect of descriptive and injunctive norms on behavior: descriptive norms have a larger effect and “affect behavior primarily directly, whereas the effect of injunctive norms relies on the indirect effect through intentions” (p. 13).
Within a TPB framework, social norms have traditionally been conceptualized as injunctive norms (i.e., “what significant others think the person ought to do”; Rivis & Sheeran, 2003, p. 219). In their meta-analysis, Armitage and Conner (2001) conclude that the social norm construct exhibits the
18
weakest relationship to intention. However, they posit that measurement issues and the way social norms are conceptualized (i.e., injunctive norms) are central in explaining the construct’s weak association with intention. A subsequent meta-analysis (Rivis & Sheeran, 2003) thus explored the additional effect of descriptive norms and demonstrated how the construct significantly contributes to explaining variance in intention, over and above the TPB main constructs. More recently, Manning (2009) investigated the effects of injunctive and descriptive norms across behaviors in a meta- analytical synthesis, yielding further support for the larger effect of descriptive norms vis-à-vis injunctive norms.
The act of eating or choosing what food to eat is susceptible to social influences. For example, Higgs (2015) explored how social norms affect eating behaviors. She posits that the presence of others has a powerful effect on behavior “because following (or not following) norms is associated with social judgements” (p. 42). Modeling, or the effect of the presence of others when eating (Herman et al., 2003; Vartanian et al., 2015), is a related concept to descriptive norms. Both concepts involve observing others’ behavior as a means of deciding what is normal conduct. Vartanian et al. (2015) conducted a meta-analysis of research on modeling of food intake, demonstrating a strong modeling effect “such that participants ate more when their companion ate more, and ate less when their
companion ate less” (p. 119). Robinson et al. (2014) reviewed studies that experimentally manipulated information about eating norms and found a consistent effect on eating behavior. They concluded that providing information suggesting that other people are eating healthily influences both the quantity and types of food people choose to consume.
Previous consumer research on functional food has explored the role of social norms on behavior. For example, O’Connor and White (2010) employed the TPB to study non-users of functional food and their willingness to consent to a free trial of an unspecified functional food which involved consuming the product every day over the next two months. Social norms (i.e., injunctive norms) were
significantly associated with intention. Another study (Patch et al., 2005), also using the TPB and injunctive norms, did not find a significant association with intention. Yet another, more recent study (Nguyen et al., 2020) demonstrates significant associations of injunctive norms and both attitude and intention toward the purchase of functional yoghurt among Vietnamese consumers. Salmani et al.
(2020) combined measures of injunctive and descriptive norms into a summary construct labeled subjective norms and found a significant association with the use of vitamin-enriched foods. Rezai et al. (2014), using constructs from both the HBM and the TPB, found that injunctive norms “have a direct influence on consumer intention to accept synthetic functional foods” (p. 30). Wang and Chu (2021) proposed a mediation model to study the influence of descriptive and injunctive norms on intention to purchase functional foods. Both descriptive and injunctive norms were indirectly related
